Reds Hackathon 2025

# install dplyr

install.packages("dplyr")

Error in install.packages : Updating loaded packages

install.packages("ggplot2")

Error in install.packages : Updating loaded packages

#load in dplyr

library(dplyr)
library(ggplot2)
library(datasets)

# read the csv files

lahman_people <- read.csv("lahman_people.csv")
savant_data <- read.csv("savant_data_2021_2023.csv")
install.packages("ggplot2")

trying URL 'https://cran.rstudio.com/bin/macosx/big-sur-arm64/contrib/4.4/ggplot2_3.5.1.tgz'
Content type 'application/x-gzip' length 4974305 bytes (4.7 MB)
==================================================
downloaded 4.7 MB

The downloaded binary packages are in
    /var/folders/zh/vh4pt2mx1z56qx417_jrsbp00000gn/T//RtmpaTOuiW/downloaded_packages

install.packages("dplyr")

Error in install.packages : Updating loaded packages

head(lahman_people)

head(savant_data)

install.packages("dplyr")

trying URL 'https://cran.rstudio.com/bin/macosx/big-sur-arm64/contrib/4.4/dplyr_1.1.4.tgz'
Content type 'application/x-gzip' length 1599250 bytes (1.5 MB)
==================================================
downloaded 1.5 MB

The downloaded binary packages are in
    /var/folders/zh/vh4pt2mx1z56qx417_jrsbp00000gn/T//RtmpaTOuiW/downloaded_packages

plate_appearances <- 
    # start with the savant data
    savant_data  %>%
    # we will group by batter, season, game, and at bat and preserve the 
    group_by(
        batter,
        game_year,
        game_pk,
        at_bat_number
    ) %>%
    summarise() %>%
    ungroup() %>%
    # now we have just unique batter, season, game, and at bat observations
    # but, we need to count how many of those there are each season
    # so, we will do another group by and summarise
    group_by(
        batter,
        game_year
    ) %>%
    summarise(
        # the n() function counts the number of unique observations we have
        playing_time = n()
    ) %>%
    ungroup()

`summarise()` has grouped output by 'batter', 'game_year', 'game_pk'. You can override using the `.groups` argument.`summarise()` has grouped output by 'batter'. You can override using the `.groups` argument.

plate_appearances

pa_in_year <- plate_appearances %>%
  group_by(batter) %>%
  summarise(
    pa_2021 = sum(playing_time[game_year == 2021], na.rm = TRUE), # Plate appearances for 2021
    pa_2022 = sum(playing_time[game_year == 2022], na.rm = TRUE), # Plate appearances for 2022
    pa_2023 = sum(playing_time[game_year == 2023], na.rm = TRUE)  # Plate appearances for 2023
    ) %>%
  mutate(
    pa_avg = round(rowMeans(select(., pa_2021, pa_2022, pa_2023), na.rm = TRUE)) # Calculate row-wise mean
    )
pa_in_year

NA

pa_in_year <- pa_in_year %>%
  mutate(
    # Is there a steady decrease in plate appearances?
    # Calculate percentage decrease between 2021 and 2022
    decrease_21_22 = (pa_2021 - pa_2022) / pa_2021,
    # Calculate percentage decrease between 2022 and 2023
    decrease_22_23 = (pa_2022 - pa_2023) / pa_2022,
    # Check if both decreases are at least 15%
    decreasing = if_else(
      decrease_21_22 >= 0.15 & decrease_22_23 >= 0.15,
      1,
      0,
      2 # likely got injured or underperformed (# -> 0)
    ),
    
    # Is there a steady increase in plate appearances?
    # Calculate percentage increase between 2021 and 2022
    increase_21_22 = (pa_2022 - pa_2021) / pa_2021,
    # Calculate percentage increase between 2022 and 2023
    increase_22_23 = (pa_2023 - pa_2022) / pa_2022,
    # Check if both increases are at least 15%
    increasing = if_else(
      increase_21_22 >= 0.15 & increase_22_23 >= 0.15,
      1,
      0,
      2 # likely a rookie or comeback from injury (0 -> #)
    ),
    
    # Is the amount of plate appearances constant within 15%
    constant = if_else(
      # if it is increasing or decreasing, then it is not constant
      (increasing == 1) | (decreasing == 1), 
       0,
       if_else(
         # If EITHER difference has an increase or decrease from 0%-15%, it is roughly constant
         (((increase_21_22 < 0.15) & (increase_21_22 > 0)) | 
            ((decrease_21_22 < 0.15) & (decrease_21_22 > 0))) &
           (((increase_22_23 < 0.15) & (increase_22_23 > 0)) | 
              ((decrease_22_23 < 0.15) & (decrease_22_23 > 0))), 
         # if both differences are less than a 15% change it is roughly constant
         1,
         0
         )
       )
    
  ) %>%
  # Optionally, remove intermediate columns
  select(-decrease_21_22, -decrease_22_23, -increase_21_22, -increase_22_23)
pa_in_year

plot(pa_in_year$pa_2021, pa_in_year$pa_2022,
     pch = 19,
     cex = 2,
     col = if_else(pa_in_year$decreasing == 1 | pa_in_year$decreasing == 2, 
                   "red", 
                   if_else(pa_in_year$increasing == 1 | pa_in_year$increasing == 2,
                           "#3d943c",
                           if_else(pa_in_year$constant == 1,
                                   "blue",
                                   "black"))))

plot(pa_in_year$pa_2022, pa_in_year$pa_2023,
     pch = 19,
     cex = 1.5,
     col = if_else(pa_in_year$decreasing == 1, 
                   "red", 
                   if_else(pa_in_year$increasing == 1,
                           "#3d943c",
                           if_else(pa_in_year$constant == 1,
                                   "blue",
                                   "black"))))

summary(factor(pa_in_year$decreasing))

  0   1   2 
696 125 518 

summary(factor(pa_in_year$increasing))

   0    1    2 
1112  115  112 

summary(factor(pa_in_year$constant))

   0    1 
1285   54 

batters_faced <- 
    # start with the savant data
    savant_data  %>%
    # we will group by batter, season, game, and at bat and preserve the 
    group_by(
        pitcher,
        game_year,
        game_pk,
        at_bat_number
    ) %>%
    summarise() %>%
    ungroup() %>%
    # now we have just unique batter, season, game, and at bat observations
    # but, we need to count how many of those there are each season
    # so, we will do another group by and summarise
    group_by(
        pitcher,
        game_year
    ) %>%
    summarise(
        # the n() function counts the number of unique observations we have
        playing_time = n()
    ) %>%
    ungroup()

`summarise()` has grouped output by 'pitcher', 'game_year', 'game_pk'. You can override using the `.groups` argument.`summarise()` has grouped output by 'pitcher'. You can override using the `.groups` argument.

batters_faced

plot(factor(batters_faced$game_year), batters_faced$playing_time, cex = 0.5)

bf_in_year <- batters_faced %>%
  group_by(pitcher) %>%
  summarise(
    bf_2021 = sum(playing_time[game_year == 2021], na.rm = TRUE), # batters faced for 2021
    bf_2022 = sum(playing_time[game_year == 2022], na.rm = TRUE), # batters faced for 2022
    bf_2023 = sum(playing_time[game_year == 2023], na.rm = TRUE)  # batters faced for 2023
    ) %>%
  mutate(
    bf_avg = round(rowMeans(select(., bf_2021, bf_2022, bf_2023), na.rm = TRUE)) # Calculate row-wise mean
    )
bf_in_year

bf_in_year <- bf_in_year %>%
  mutate(
    # Calculate percentage decrease between 2021 and 2022
    decrease_21_22 = (bf_2021 - bf_2022) / bf_2021,
    # Calculate percentage decrease between 2022 and 2023
    decrease_22_23 = (bf_2022 - bf_2023) / bf_2022,
    # Check if both decreases are at least 15%
    decreasing = if_else(
      decrease_21_22 >= 0.15 & decrease_22_23 >= 0.15,
      1,
      0,
      2 # likely injured or under-performed (# -> 0)
    ),
    
    # Calculate percentage increase between 2021 and 2022
    increase_21_22 = (bf_2022 - bf_2021) / bf_2021,
    # Calculate percentage increase between 2022 and 2023
    increase_22_23 = (bf_2023 - bf_2022) / bf_2022,
    # Check if both increases are at least 15%
    increasing = if_else(
      increase_21_22 >= 0.15 & increase_22_23 >= 0.15,
      1,
      0,
      2 # likely rookie or comeback player (0 -> #)
    ),
    
    # Is the amount of batters faced constant within 15%
    constant = if_else(
      # if it is increasing or decreasing, then it is not constant
      (increasing == 1) | (decreasing == 1), 
       0,
       if_else(
         # If EITHER difference has an increase or decrease from 0%-15%, it is roughly constant
         (((increase_21_22 < 0.15) & (increase_21_22 > 0)) | 
            ((decrease_21_22 < 0.15) & (decrease_21_22 > 0))) &
           (((increase_22_23 < 0.15) & (increase_22_23 > 0)) | 
              ((decrease_22_23 < 0.15) & (decrease_22_23 > 0))), 
         # if both differences are less than a 15% change it is roughly constant
         1,
         0
         )
       )
    
  ) %>%
  # Optionally, remove intermediate columns
  select(-decrease_21_22, -decrease_22_23, -increase_21_22, -increase_22_23)
bf_in_year

plot(bf_in_year$bf_2021, bf_in_year$bf_2022,
     pch = 19,
     cex = 2,
     col = if_else(bf_in_year$decreasing == 1 | bf_in_year$decreasing == 2, 
                   "red", 
                   if_else(bf_in_year$increasing == 1 | bf_in_year$increasing == 2,
                           "#3d943c",
                           if_else(bf_in_year$constant == 1,
                                   "blue",
                                   "black"))))

plot(bf_in_year$bf_2022, bf_in_year$bf_2023,
     pch = 19,
     cex = 2,
     col = if_else(bf_in_year$decreasing == 1, 
                   "red", 
                   if_else(bf_in_year$increasing == 1,
                           "#3d943c",
                           if_else(bf_in_year$constant == 1,
                                   "blue",
                                   "black"))))

summary(factor(bf_in_year$decreasing))

  0   1   2 
993 137 253 

summary(factor(bf_in_year$increasing))

   0    1    2 
1026  142  215 

summary(factor(bf_in_year$constant))

   0    1 
1331   52